Clarify the consequences. Different outcomes will have different consequences, and these, too, must be defined. In general, you should follow the same process for defining consequences as we laid out in Chapter 5, expressing them as precisely as necessary to make an informed choice. Depending on the complexity of the decision, you should lay out the consequences in one of three ways:
•A written description. Although the least precise, a broad written description may occasionally be good enough. But remember that, whereas phrases such as ‘‘marginal,’’ ‘‘OK,’’ or ‘‘a waste of effort with little to show’’ may suit personal decisions, they require too much interpretation to be readily communicable to others.
•A qualitative description by objective. Consequences expressed qualitatively by objective include more information than simple written descriptions, as they break a consequence into its constituent parts. For an outdoor picnic, the consequences of sunny weather for each of Janet’s objectives would be described as (1) high on fun, (2) high on family involvement, and (3) low on cost.
•A quantitative description by objective. Though they may require the most time to develop, consequences expressed quantitatively by objective—such as cost estimates in dollars— are the clearest, the most easily comparable, and the easiest to use. The cost of a used car listed as ‘‘$5,000 plus or minus 10 percent’’ is more useful and meaningful than one listed as ‘‘low.’’
In all cases, though, keep in mind that descriptions of consequences need only be precise enough to provide the information needed to reach a smart choice. If your choice is clear with a written description, there’s no need to spend the time to develop precise, quantitative estimates.
Picture Risk Profiles with Decision Trees
Often, the development of risk profiles can itself clarify uncertainty to the point where the smart choice becomes obvious. But not always. Some decisions, particularly highly complex ones, will require further analysis. That’s when a decision tree can be extremely useful. A decision tree provides a graphical representation—a picture—of the essence of a decision, displaying all the interrelationships among choices and uncertainties. In one sense, a decision tree is like a blueprint—it lays out, methodically and objectively, the architecture of a decision. And just as a builder would not set out to construct a house without a blueprint, a decision maker will often require a decision tree to resolve a tough choice under uncertain conditions.
The essence of Janet Ellingwood’s employee party problem, for example, can be plotted in a decision tree, as we see on page 124. The tree begins at the point of the decision (the square labeled 1), with the initial branches representing the competing alternatives. In Janet’s case, there are two alternatives, hotel dinner dance and mountain picnic, so there are two branches. Each alternative branch leads to a fork (the circles labeled 2 and 3), indicating an uncertainty. Each possible outcome of the uncertainty—in this case, rain or shine—is represented by a branch leading out from the fork. These outcome branches are labeled with their respective chance of occurring. (Janet uses 30 percent for the chance of rain, a judgment she elicited from a local meteorologist.) Each of the outcome branches in turn leads to different consequences, which are summarized, by objective, at the tips of the tree.
Decision Tree for Janet’s Employee Party
* * *
This simple decision tree, with its four possible paths, shows how pictures can clarify the relationships among alternatives, uncertainties, and consequences. It brings risk profiles to life. Seeing her decision presented this way immediately sharpens Janet’s thinking. She concludes that a successful picnic would meet her objectives so much better than would the dinner dance that it is worth taking a 30 percent chance on rain. She opts for the picnic.
Decision trees are especially useful for explaining decision processes to others. (Hence the careful numbering of the branching points and the labeling of the branches.) Getting into the habit of sketching decision trees, even for relatively simple decisions involving uncertainty, can enhance your decision-making skills in two ways. First, decision trees encourage thorough, logical thinking about a problem—a useful habit to cultivate. Second, mastering the mechanical skill of tree construction on simple problems will make it easier to use the technique for more complex ones, such as the one illustrated in the following application.
APPLICATION
To Settle or Not to Settle?
Karen Plavonic hasn’t had a good night’s sleep in weeks. Her stomach is always in knots. Day and night she anguishes over whether to accept a $300,000 offer to settle her personal injury lawsuit. On the one hand, she knows there’s a good chance of getting much more— maybe as much as a million dollars—if she refuses and goes to trial. But, on the other hand, she could lose in court and end up with nothing. Then she’d wish she had accepted the offer (and she knows her mother would never let her forget her mistake!).
Karen, 27 years old and single, feels she may have contributed in a small way to the automobile accident that has left her slightly disabled, disfigured, and plagued with mounting medical expenses. Though she doesn’t want to look foolish for ‘‘throwing away’’ the settlement, her lawyer, Sam Barnes, is pressuring her in the opposite direction. He is urging her not to weaken, not to let the other guy off the hook. Karen, however, can’t overcome residual feelings of guilt about the accident, despite her relative innocence and the greater harm she has suffered — facial scars, impaired mobility in her neck and left shoulder, and loss of income. She feels keenly the possibility that she might break down in court, jeopardizing her case. Every friend, relative, coworker, and acquaintance she’s ever had is giving her conflicting advice about what to do. She just can’t decide.
Karen’s Decision Problem
Karen’s close friend Jane Stewart has suffered with her through the aftermath of the accident, and she is now serving as Karen’s sounding board for her soul-wrenching inner debate about the lawsuit. Jane, a management consultant with professional experience in facilitating decision making, has undertaken to help Karen think through her situation systematically, to end her emotionally devastating indecision. Jane wants to help Karen decide whether she should go to court or settle out of court and to feel comfortable that she is making an appropriate decision. As she tells Karen, who hopes her luck will finally turn, ‘‘Most of the time luck favors the better decision maker.’’
Together, Karen and Jane isolate three essential considerations on which Karen’s decision will hinge:
1.The chance of winning the trial and, if won, the chances of different possible jury awards.
2.The time and psychological stresses associated with going to trial and of not going to trial, together with the degree of Karen’s regret if she loses or elation if she wins.
3.Karen’s willingness to take risk.
In addition to going to court or settling for $300,000, Karen and Jane recognize a third alternative: waiting for a better settlement offer. Based on his knowledge of the opposing lawyer, Sam doesn’t think another offer will be forthcoming. But Karen and Jane decide that, if Karen chooses to settle, she should keep her options open until the last minute.
To complete her risk profile of the go-to-court alternative, Karen will need to hear Sam’s judgment about the chance of a positive trial outcome and of different award amounts. Karen schedules a meeting with him and Jane, for which Jane prepares some materials.
Karen’s Decision Tree
At their meeting, Jane passes around a diagram (page 128) that describes Karen’s decision problem as a decision tree. Reading from the left, the box labeled 1 represents Karen’s basic decision: go to court or settle out of court. Deciding to settle, the lower branch, entails no uncertainty. But the decision to go to court, the upper branch, leads to two uncertainties: Will Karen win or lose (fork 2), and if she wins, how much will she get (fork 3)?
The range of figures off fork 3, from $200,000 to $1,000,000, represents the possible jury awards
, which Jane derived from Karen’s earlier discussions with Sam. The figure $210,000 at the end of the settlement branch represents what Karen would have left after paying Sam his 30 percent fee. In addition to money, the tree notes two other possible consequences: ‘‘psych’’ indicates that the outcome might exact nonmonetary costs, including sleeplessness, anxiety, and regret, and ‘‘time’’ indicates that the outcome would entail a further investment of time.
Karen’s Decision Tree
* * *
Karen’s Chances
Karen and Jane now call on Sam’s expertise to quantify the likelihood that Karen will win the trial. Sam has told Karen that she has a ‘‘pretty good’’ chance of winning, based on the outcomes of similar cases, the record of the judge, and his assessment of his own litigation skills.
Jane probes the meaning of ‘‘pretty good,’’ trying to arrive at a hard number, which would sharpen the analysis. She asks Sam, ‘‘How would you translate ‘pretty good’ into a probability?’’
‘‘I just don’t think that way,’’ Sam answers. ‘‘I don’t see how you can put a number on everything, especially things as subjective as winning a trial.’’
Jane turns to Karen. ‘‘How do you interpret that, Karen? Give me some number.’’
‘‘Oh, I’d say that Sam thinks our chance of winning is around 20 or 30 percent.’’
Sam protests. ‘‘That’s not what I said! When I say a pretty good chance, I mean something more than that!’’
‘‘How much more? More than 50-50?’’
‘‘Certainly. More than 50 percent.’’
‘‘How much more?’’
‘‘Oh, I don’t know that you can put a precise number on it. It certainly isn’t as high as 90 percent. In jury trials you can never be that sure. It’s maybe somewhere between 60 and 80 percent.’’
‘‘Would you say that 70 percent is reasonable, or high, or low?’’
‘‘It’s a good estimate, as close as we can get.’’
‘‘OK, let’s talk about the uncertainties of the jury award at fork 3.’’
Jane probes Sam’s knowledge about the possible jury award. After an hour or so of give and take, she prepares a table (below) summarizing his judgments. It divides the previously determined $800,000 range ($200,000 to $1 million) into four equally likely outcome intervals, labeled ‘‘Low,’’ ‘‘Medium,’’ ‘‘High,’’ and ‘‘Very high.’’ A representative amount for each interval, also listed in the table, makes it easier to interpret the implications of the jury award uncertainty. The figure $300,000, for example, stands for the range $200,000 to $410,000.
Chances for the Jury Award if Karen Wins
* * *
Karen’s Consequences
The task now is to factor in the nonmonetary costs to Karen of pursuing a jury trial: her time, her anxieties about losing, her lingering guilt about her role in the accident, her apprehension of the criticism of others (especially her mother) if she does lose after turning down a sure thing, and her own potential regrets on the same score.
Karen and Jane use the even swap method (as described in Chapter 6) to assign monetary values to the intangible costs. As shown in the ‘‘Adjustments’’ column for summarizing Karen’s consequences (below), they use negative dollar amounts to represent what Karen would sacrifice, or ‘‘pay,’’ to rid herself of all time and psychological impacts, and positive dollar amounts to represent the equivalent value she would gain, or ‘‘earn,’’ from her elation at winning a very high award. The adjustment figures vary with the amount of the award, reflecting different balances of anxiety to satisfaction or regret to elation.
Net Equivalent Dollar Consequences for Karen
* * *
Once they have established equivalent monetary values of the intangible costs, Karen and Jane can add those values to (or subtract them from) the award amount to calculate the overall value of each outcome. The numbers in the column labeled ‘‘Equivalent Monetary Amount’’ thus represent the net values to Karen from the representative awards, after deducting her lawyer’s fees (30%) and factoring in the adjustment amounts. Jane adds the net values as well as the chances to Karen’s decision tree (see below). She also summarizes them in a risk profile (see page 132), where the 17.5 percents represent the 70 percent chance of winning the trial times the 25 percent of each award.
Karen’s Decision Tree with Consequences and Probabilities Added
* * *
Karen’s Risk Profile for Going to Court
* * *
Karen sighs. ‘‘You sure have clarified what’s at stake in my decision, Jane. But I still don’t know what I should do. Should I take the $300,000? Or should I take my chance in court?’’
‘‘Well,’’ Jane responds, ‘‘that depends on how you feel about taking risks. That’s the final piece of the puzzle.’’
(To be continued in Chapter 8.)
Lessons from the Application
Thanks in large part to Jane’s guidance and Sam’s input, Karen now has excellent risk profiles for her two alternatives: to settle or to go to court. Her case illustrates four essential points to keep in mind when describing and comparing risk profiles.
•Strive to use numbers to clarify the chances of different outcomes. People sometimes take refuge in vague qualitative descriptions of chances in order to avoid commitment, responsibility, or second-guessing. They may have to be pressed hard to quantify their judgments, but as Karen’s case demonstrates, the greater precision and usefulness of numbers makes it worth the effort needed to get them.
•Clarify the consequences by being specific. By dividing the broad span between $200,000 and $1,000,000 into four narrower, equally likely ranges, with a representative dollar amount for each, Karen gained a much better appreciation for what winning might mean to her.
•Use the even swap method to convert intangible concerns into a meaningful equivalent value. This process improved Karen’s understanding of the possible consequences of her choices, because it helped her to think hard about how much she valued the ‘‘intangibles.’’ She could then combine this with any dollar award she might receive, subtract attorney’s fees, and arrive at a single indicator of the net equivalent dollar amount for that consequence.
•Take time to think about the important uncertainties influencing a decision. Constructing risk profiles does not require much time or effort or any specialized knowledge. It does require an honest effort to identify the key uncertainties and their possible outcomes and to clarify the chances and consequences of each.
CHAPTER 8
Risk Tolerance
A PERSON’S ATTITUDE TOWARD RISK is as individual as his or her personality. Some people avoid risk at all costs—they put all their retirement savings into certificates of deposit insured by the federal government. Others embrace risk—they invest all their money in options, in penny stocks, in junk bonds. Most of us fall somewhere in between. We take on some degree of risk, knowing that it goes hand in hand with reward, but not so much that we can’t sleep at night.
How can you take your personal tolerance for risk into account in making decisions? We saw in the last chapter that choosing under uncertainty boils down to choosing among the risk profiles of your various alternatives. Once you’ve specified the risk profiles, you can compare them and readily eliminate all the poor choices. Often, the best choice will be obvious. But suppose you’ve done this and you still can’t make up your mind. At this point, you need to focus not just on the risk profile, but on the degree of risk you are willing to assume.
Consider, for example, the following dilemma. Years ago, Robert Goldman, now 68, lost the vision in his left eye. His vision in his right eye has deteriorated in recent years due to a cataract. Now, even with his glasses, his corrected vision is only about 20/50, and anything he looks at is fuzzy, particularly at night. As a result, he has been advised to stop driving his car after dark.
Recently, Rob had his eyes examined, and his doc
tor, Joycelynn Eddy, raised the possibility of cataract surgery. For his particular case, the chances are 90 percent that cataract surgery would be successful, ‘‘success’’ meaning that his vision would be restored to 20/30, corrected, with no fuzziness. An unsuccessful outcome, which Dr. Eddy estimates as having a 10 percent likelihood, would erode his vision to no better than 20/100, corrected, with persisting fuzziness.
Rob quickly draws a decision tree (below), documenting the two alternatives (surgery or no surgery) and the two possible outcomes (successful or unsuccessful). The consequences of each outcome are described in terms of two fundamental objectives: acuity and clarity. The tree clearly shows the risk profiles of the choices, yet Rob finds that the decision remains quite difficult. He would love to eliminate the fuzziness and once again have near-normal vision—and the chances of this are very good. But if the surgery failed, Rob would be worse off than he is now. He would have to quit driving altogether, give up some of his everyday physical activities, and curtail his reading or invest in expensive enlargers and other aids.
Rob’s Cataract Surgery Decision Tree
* * *
The essence of Rob’s decision is that the surgery alternative offers him a 90 percent chance of restoring his vision, but a 10 percent chance of permanently worsening it. Clear enough? Yes. And yet so difficult. Should he take his chances on the surgery or play it safe with the status quo?
Smart Choices Page 11
